"Everyone is doing it. Projects older than me, and developers who could be seen as role models and important figures in the space are adopting LLMs into their development workflow more readily than anything I'd ever seen. Vim, VLC, gstreamer, Kitty, the Linux Kernel, these are all already actively integrating the lying machine into their workflows.
Am I just being a vegan about it?"
I would argue that we who refuse to use LLMs are more of "AI" antifascists than "AI" vegans. Only one third of voters voted for the nazis. A certain percentage opposed them. But many of those who didn't actively support them turned a blind eye, because they didn't want to make waves, the system benefited them, or at least they had comfortable enough lives under it. It is the same with technofascism*. Until it stops being comfortable/cheap. And then, everyone will always have been against it.
*The percentage might be closer to those of vegans though.
(HT @pelle )
@rhelune @pelle From the link, "Software development is primarily about coming up with solutions that work, a product, an artefact, something your user can run and have it work. And, well, the chat bot can do that just fine."
This is false. A chat bot can reproduce known algorithms (or tbf copies of whole programs) , but software development requires data as well. The data modeling aspects are beypnd the abilities of LLMs.
@lwriemen @pelle LLMs can only solve problems that have been solved many times before. The question is, how much of software development is solving new problems at all.